Major knee surgery is associated with moderate or severe post-operative pain. Intrathecal morphine and continuous femoral 3-in-1 block were compared prospectively in 40 patients for pain after major knee surgery under spinal anaesthesia, with 4 mL isobaric 0.5% bupivacaine. In a random order, 20 patients received preservative free morphine 0.3 mg mixed with spinal bupivacaine. In 20 patients, following spinal anaesthesia with only bupivacaine, femoral 3-in-1 block was performed post-operatively with 0.5% bupivacaine 2 mg kg-1. The block was continued via a catheter with 0.25% bupivacaine 0.1 mL h-1 kg-1 until the next morning (24 h after induction of spinal anaesthesia). Intramuscular oxycodone was given as a rescue analgesic in all patients. Two patients from the femoral group were excluded due to technical failure. Three patients in the morphine group and one patient in the femoral group did not need any additional oxycodone. In the morphine group on average 2.8 (range 0-7) and in the femoral group 3.2 (0-5) additional doses of oxycodone were needed during the 24 h observation period. The mean pain scores were significantly lower in the morphine group at 9 and 12 h into the 24-h trial. Itching was seen only in the morphine group (40% of the patients). Other side effects were similar in the two groups. All patients were satisfied with their pain therapy. Both intrathecal morphine and femoral 3-in-1 block alone were insufficient for the treatment of severe pain after major knee surgery.